Borbely DC 100 Amp: Left Channel Troubleshooting

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
An acquaintance asked me to look at this DC 100 he built back in the day. The left channel is out. At power on, the right channel heat sink warms up nice and toasty, but the left is cold. Here are the voltages I'm seeing, left and right, for the K135 and J50 output transistors:


LEFT
K135 pair

Drain: 57.8V and 57.4V
Gate: 0.275V and 0.264V
Source: .003V and .001V


RIGHT
K135 pair
Drain: 57.0V and 56.9V
Gate: 1.1V and 1.09V
Source: .048V and .05V


LEFT
J50 pair
Source: .004V and .003V
Gate: -.6V and -1.1V
Drain: -57.5V and -57.5V


RIGHT
J50 pair
Source: .059V and .050V

Gate: 1.129V and 1.11V

Drain: 57.6V and 57.2


I thought the issue might lie with the bias circuit, so I replaced Q16 and Q17 (K216 and J79 respectively), the zener diodes, the 1N4148s, and Q13 (MPS-A05). No change. When I compare the voltages of Q13, left and right, here's what I get:



LEFT
Q13 (MPS-A05)
Collector: 1.6V
Base: 1.0V
Emitter: 0.9V


RIGHT

Q13 (MPS-A05)
Collector: 1.1V
Base: 0.4V
Emitter: 1.8V


I'm going to continue checking the various components on the board, but relative novice at troubleshooting that I am, I welcome any suggestions.
 
Administrator
Joined 2007
Paid Member
We would have to see the circuit details to be able to advise on this. Although your measurements seem to indicate a lack of bias, that alone would normally only give rise to a small increase distortion so it sounds like something else is going on.

As always... a circuit diagram is a must.
 
The bases of Q13 should be approx 0.7V more positive than their emitters . None of the voltage readings around Q13 either left or right channel instill any confidence in their accuracy . Have you identified the leads correctly . Trying to work with the supplied voltages have you checked the resistance/ continuity of all the components associated with Q13
 
Administrator
Joined 2007
Paid Member
Thanks...

Quick thoughts and things to check:

1/ That you have correct positive supply voltage on each end of R34 (not easy to make out some numbers) and also the correct negative voltage on the similar 22 ohm on the negative rail below it.

2/ Check the 5 volt references are correct (T7 and T10) and make sure you have approx. 4.4 volts across each 560 ohm (next to T7 and T10 on diagram).

3/ Check the voltage across the two zeners (marked D3, D6 and D7 and D10). The diode voltage isn't marked on the diagram so you will have to compare with the good channel. All four should be similar.

4/ The bias generator Q13 can only really give a fault of excess current or no current in the output stage. High current would be destructive, low current would not stop the amp from working to the point that you may not even notice.

One reading that can be inferred but needs checking is that voltage on L1 is zero. The driver voltages suggest it will be but you should check.
 
Thanks...

Quick thoughts and things to check:

1/ That you have correct positive supply voltage on each end of R34 (not easy to make out some numbers) and also the correct negative voltage on the similar 22 ohm on the negative rail below it.

2/ Check the 5 volt references are correct (T7 and T10) and make sure you have approx. 4.4 volts across each 560 ohm (next to T7 and T10 on diagram).

3/ Check the voltage across the two zeners (marked D3, D6 and D7 and D10). The diode voltage isn't marked on the diagram so you will have to compare with the good channel. All four should be similar.

4/ The bias generator Q13 can only really give a fault of excess current or no current in the output stage. High current would be destructive, low current would not stop the amp from working to the point that you may not even notice.

One reading that can be inferred but needs checking is that voltage on L1 is zero. The driver voltages suggest it will be but you should check.




Thanks Mooly. Here's what I came up with.



1. Supply voltages
LEFT
R34: 59.9V and 59.2V

R41: -59.9V and -59.2V


RIGHT

R34: 57.8V and 57.2V
R41: -57.9V and -57.2V





2. Reference voltages
LEFT

T7: 53.8V and 58.8V


RIGHT
T7: 57.6V and 52.6V




LEFT
T10: -54.1V and -59.1V


RIGHT
T10: -52.8V and -57.8V


I have 5V across each 560 ohm resistor next to T7 and T10.




3. Zener voltages
LEFT
D3: 19.4V and 14.5V
D6: 4.8V and 0
D7: -19.1V and -14.3V
D10: -4.7V and 0


RIGHT
D3: 19.1V and 14.3V
D6: 4.8V and 0
D7: -19.0 and -14.3V
D10: -4.7V and 0


4. I did in fact err in my measurements of T13. :eek: They should be as follows.
LEFT
E: 0.9V
B: -1.0V
C: -1.6v


RIGHT
E: 1.8V
B: -0.4V
C: -1.1v


As for L1, yes, the voltage is zero.
 
The voltage at the top of R38, left side, is +0.25V, and the bottom is -1.11. For the right, it's +1.12 on top and -0.67 on the bottom.

There are four final output transistors (J50 and K135) per channel. I'm not sure what "lateral mosfets" means. How they're mounted with respect to each other?
 
Administrator
Joined 2007
Paid Member
Those are lateral FET's. The best :)

It refers to the type and construction of the semiconductor material. They also have a very low turn on voltage threshold compared to the more common vertical FET used in many of todays amps.

A bias voltage of around 800 mv between the two gates of the output devices would be enough to get them conducting, higher voltage would get them conducting more... and you seem to have that on both channels.

Without knowing all the circuit details of the output stage this may sound a bit daft but can you check that there is supply voltage on the output devices. Are they fused separately ? Its worth checking because the DC conditions in this part of the circuit seem OK to me. Nothing obvious has shown up yet.

Check the supplies to the outputs. Look for rail fuses to the output transistors and also fuses on the output line of the amp. Also see if there are any low value resistors around the output FET's (around 0.22ohm) and if so are they OK ? That bias voltage (assuming a standard output stage) should be getting it to conduct and so heat up... and its not. We need to find out why. So check those supplies on the outputs themselves.

I'll look in again tomorrow.
 
What we need is a lifted pcb pad at R39 which both breaks the feedback rail /output and the continuity between the upper and lower mosfets .
 

Attachments

  • dc100_outputs.jpg
    dc100_outputs.jpg
    263.5 KB · Views: 241
The supply voltages at the fuses, to the PCB, and then to the output transistors seem fine: They're +59V -59V, slightly higher than the +58V -58V of the right.

I meant to post the circuit details of the output stage, but epicyclic beat me to it. See his post with the attachment.

There's no PCB for this stage, but I checked the solder connection at R39 and confirmed continuity in all directions. As for the 0.22 ohm resistors, I can desolder a leg on each to check their resistance, but unfortunately none of my meters provide useful resistance measurements down that low.

Fwiw (which probably isn't much), here are some pictures of the patient.
 

Attachments

  • 20180616_235312.jpg
    20180616_235312.jpg
    552.5 KB · Views: 227
  • 20180616_235041.jpg
    20180616_235041.jpg
    508.8 KB · Views: 222
  • 20180616_235153.jpg
    20180616_235153.jpg
    260.2 KB · Views: 222
  • 20180616_235254.jpg
    20180616_235254.jpg
    444.5 KB · Views: 221
Administrator
Joined 2007
Paid Member
The drawing is a bit unconventional (drawing a connecting link through the FET symbol) as the FET's are actually in parallel, each having a 0.22 ohm in each FET's source.

Lets go back to basics. The voltage across the 150 ohm at the left (R34 in the latest diagram) is what will determine if the FET's should be drawing current.

Can you recheck the voltage across these 150 ohm resistors. It is a single reading we are after (red lead on the top of the resistor, black lead on the bottom end). If you have 800mv (approximate, probably even less than that in practice) then the FET's should be on and pulling current. So compare both channels. Lets see what is really across the FET gates and whether they should be on.

The 0.22 ohms can be checked in circuit, if they read essentially short circuit then they are going to be OK.

Lets just clear up this bias question first.
 
Not sure how much of a difference it makes (I suspect some), but the part of the diagram where R34 is shown refers to the Servo 100 amp. Only the part within the bounds of the dashed line marked HEATSINK ASSEMBLY is for the DC 100 (the amp I have). We have to refer back to the other circuit diagram for the DC 100 driver circuit. Looking at that, I believe you want a measurement for the 150 ohm resistor R38. In that case, I get this:



R38

LEFT: +1.424V
RIGHT: +1.790V


For the 0.22 ohm resistors, my meter shows 0.3 for all of them, so they seem fine.
 
Administrator
Joined 2007
Paid Member
Interesting, and yes, it is across the 150 ohm that was of interest.

So 1.4 volts should definitely be turning the FET's on quite significantly. That voltage should also vary with adjustment of the bias preset.

Although not 100% accurate, a quick simulation of that bias voltage across those lateral FET's should be letting around 280ma flow... split between each pair. That in turn should certainly cause them to get a little warm.

We need to move past subjectively feeling the temperature of them and actually measure the current flowing. That's dead easy by simply measuring the voltage across the 0.22 ohms (all four should be reasonably similar). The voltage will be low, around 22 millivolts would be equal to 100ma flowing.

I think we need to do that to make sure that they really are conducting as they should.
 
Administrator
Joined 2007
Paid Member
Something else to check (again one of those daft things).

Does the final design of the amp use any kind of speaker relay ? These would be for DC offset protection and also add a delay to give a totally silent switch on condition.

There doesn't seem to be from what I can tell, but a home build may have differences or may have had one added.

Which leads on to the next question... have you continuity from the speaker output of the left channel all the way back to the amplifiers main output (which is the junction of the four 0.22 ohms). Any hidden speaker fuse anywhere ?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.